Slewmaster - CFA vs. VFA "Rumble"

I've run into my next glitch with my Slewmonsters. Bias seems to be dropping too much as temperature rises. I set bias at 60mA at startup. I ran it for half an hour and warmed up the whole heat sink. The devices were at 155F Bias had dropped to 25mA. I readjusted bias to 60mA and let it play under lighter load. Bias was 105mA by the time the output devices were at 110F. Once warmed up the drivers seem to be staying around 115F. I'm using BD139 for Q103 and Q104. Should I switch them to MJE340 or is there another device that will work better with the MT200s?


MJE340 .... all mine have this Vbe . 3 different outputs (4 now) all "stuck"
at 15mv (@65ma).

OS
 
Hi Jeff, you've got 13mA with original 3.9K values, which is good (they dissipate close to 0.7W in these conditions). Some 10mA out of those 13 go via zeners. Giving them more will be just "heating the zeners".
Using lower power zeners may slightly increase the voltage, but in fact, 11.8V sounds enough to me - just leave it.
OS - do you agree?


I have some IPS's that do 11.8V with smaller 4.7K's (that get warm).

Some others do 12+ with 3.9K 2w dropping resistors. I'm using the
"big" zeners - 2W ??
I have all the dropping resistors "floating" above the PCB as they all
get pretty warm.

You have to give the zeners enough power , as both the input pair ccs/or
mirror (and the servo) needs that 12V supply (or 11.8v).
Edit , actually servo takes 5ma and input pair takes 1.2ma on the "C" = 6ma burned
by the zener.
OS
 
Last edited:
Hi guys,

I have seen (fixed it by cutting track) a PCB layout error on the Kypton C PCB´s I´m building on now.
Please see picture marked with red - the track from R6 short circuit C9
An externally hosted image should be here but it was not working when we last tested it.

That error isn't there on my boards. The trace is about .020" away. Something happened at your files or the board house changed something.
 
That error would render the servo inoperable.
So I know Jwilhelm's boards do not have it.

Just reversed D8 (blue LED) , and R6 (servo return) as 220R.
Both of mine are perfect like this.

Oh also , I run R22/27 at 68R .... with led cascode trimmer R24
set at about 18K - main VAS cascode current is 5.45ma.

PS - I'm also using 22R base-stoppers for Q9/10 (R23/26).
Edit - I also "ditched" R8 on mine. (servo has full control).
OS
 
Last edited:
That error would render the servo inoperable.
So I know Jwilhelm's boards do not have it.

Just reversed D8 (blue LED) , and R6 (servo return) as 220R.
Both of mine are perfect like this.

Oh also , I run R22/27 at 68R .... with led cascode trimmer R24
set at about 18K - main VAS cascode current is 5.45ma.

PS - I'm also using 22R base-stoppers for Q9/10 (R23/26).
Edit - I also "ditched" R8 on mine. (servo has full control).
OS
Hi OS will be practical to post a new schematic with all these corrections instead of so many words
 
Hi OS will be practical to post a new schematic with all these corrections instead of so many words

Circuit will work with R8.

The R6 was a screen error and the present schema is correct.

The R22/27 change is dependent on supply voltage and device Hfe.
Edit - 47R actually is best for rail < 40V.
It can be 47- .... to even 68R+ . This would have to be determined
by operation and might not be all inclusive.

The 15R (R23/26) is also ideal , but 22,27,even 33R would work
as stopper.

To post every "tweak" on a baseline design might create more questions
than answers.
You will have to do the math , like me. (R22 I - R25 I = final VAS I)
The slew OPS will do the greater 6+ma VAS I at R22-47R , but
for easy change (with another IPS) - 5.3ma is best.

Edit - another example ...R6 can be 220R - all the way to 1K. Servo
would have to create a higher voltage to zero amp with 1K.
Very wide choices on some values !!

Edit 2 - I MIGHT add the VAS current calculation as a note on all
Hawksford or "super VAS's" .

Edit 3 - you might also note , the R22 I will increase with temp , but the
LED current will also increase , offsetting the rising total current.Output I remains
constant - cool !!
OS
 
Last edited:
Very good Idea - better documentation.

Perhaps someone good at PDF creation could take my primitive text
files and photo's and create "mini manuals" for each IPS or OPS
as I find all the tweaks and small "errata's" with them.

The first will be my present "spookyamp" and 5pair OPS.
Mine are now "tweaked out" to perfection.

OPS thermal is VERY good.
P- 14.4 - 14.7 - 15.2 - 14.4 - 14.4 (mv Re)

N - 14.2 - 14.6 - 14.9 - 14.9 - 14.5 (mv Re)

mV will become more consistent as main heatsink temp
rises. The average of 14.7mV stays rock solid over
a huge thermal range.

Spooky amp - the same 5.35ma VAS I - 0 - 40C . settles in
<1 sec. at turn-on for less than .3mv total offset -
LAB amps !

Just an example of how good it can be when a "baby"
comes home !! 😀

OS
 
Perhaps someone good at PDF creation could take my primitive text
files and photo's and create "mini manuals" for each IPS or OPS
as I find all the tweaks and small "errata's" with them.

The first will be my present "spookyamp" and 5pair OPS.
Mine are now "tweaked out" to perfection.

OPS thermal is VERY good.
P- 14.4 - 14.7 - 15.2 - 14.4 - 14.4 (mv Re)

N - 14.2 - 14.6 - 14.9 - 14.9 - 14.5 (mv Re)

mV will become more consistent as main heatsink temp
rises. The average of 14.7mV stays rock solid over
a huge thermal range.

Spooky amp - the same 5.35ma VAS I - 0 - 40C . settles in
<1 sec. at turn-on for less than .3mv total offset -
LAB amps !

Just an example of how good it can be when a "baby"
comes home !! 😀

OS
A manual for each input with setup instructions would be great. I've reread the whole thread a few times. There aren't any setup instructions for most of the inputs.